Servo Wiring and Tuning detailed How To example Mesa 7i77.

Check to see if the input pins "axis.N.motor-pos-fb" for each axis are connected to anything. (for 2.7 and older for 2.8+ the pin name may be joint.N.motor-pos-fb.

Those pins need to be connected to the encoder position feed back pins (via some signal name). That same signal name will also need to be connected to the position feedback for the PID component as well.

Thank you Todd, very well explained.
I think there is also an encoder_fb line somewhere, but not sure.
Sorry for the wague replies lately but am on phone.
Regards,
Tom.

Tommy,

No worries!

I have attached my HAL and INI if anyone wanted to take a look.

Im going to spend some quality time with the HAL manual tonight so I can have a more educated conversation here.

:woohoo:
Why are stepgens in your hal file ?????
Oh well, that explains a lot ! That is for 7i76E !
What Mesa card are you using ?
New config for servo ?
Regards,
Tom.

Tommy,
I'm running a 7i76e with a 7i85 for my encoder inputs.

My drives are pulse/direction I'm assuming that's where are the step gens coming to play??

I'm honestly not sure my hall file has been mostly untouched other than making the changes to get the 7i85 communicating.

Oh good, in that case the feedback has to come from encoders as someone mentioned here a while back.
Look for stepper motors with encoders here on the forum, there are some configs to look at to get an idea on how it works and what has to be changed. Basicaly everything with "fb" in it has to be wired to encoders in hal.
That means some 3 or 4 lines have to be changed for each axis.

I have a machine set up with step/dir servos and encoder feedback via a 7i85s (and a 5i25 FPGA). To take your original open loop step/dir config to closed loop you need to do a few modifications to the hal file.you will need to disconnect the position feedback signals from the hardware stepgen, Originally it was connected to both the motor-pos-fb and the pid.N.feedback. Switching just the motor-pos-fb from the stepgen's position feedback to the encoder's will switch the DRO and f-error, but to truly close the position loop in Linuxcnc you must switch the PID's feedback as well, and then tune the PID. You can not keep the original P=1000 setting, it will be far to aggressive a P setting for a real mechanical system. Read this as bad things will happen, start with a low P of 10 or 1.

Todd and Tommy YOU GUYS ARE GREAT!

Just had to be said, I really appreciate all the time all the board members here have taken to answer all of my questions.

Anyhow that makes perfect sense now (though you had me a bit scared that I was losing it there for a second Tommy lol )

I will give this a whirl when I get home and see how it plays out.

One thing that needs to be correct before you change over to encoder feedback is that the encoder direction and scaling must match the step/dir
scaling and direction.


If the direction is wrong, you will get an immediate runaway. If the scaling is wrong you will be unable to tune for decent performance (because FF1=1 will not be correct)

Thanks PCW i'll make sure to check that.

Before I go to much further here, my drives have an additional option for "Speed control mode" which uses +/-10v analog. Would this be a better option for my mill?

The current mode is "position mode" using step/dir

Just want to make sure I go the the best way possible.